Autoethnographic Reflexivity in Collaboration with GenAI: the co-construction of identities and knowledge

Scholarship on applications of generative AI (GenAI) tools in qualitative research has been emerging in areas such as facilitating idea generation and research design, improving content and structuring, supporting literature review, and enhancing data management and analysis, editing, reviewing, and publishing. However, the challenges in balancing the use of AI with human insight remain. An area that calls for more research is the application of GenAI tools in qualitative data analysis, particularly in autoethnographic reflections.

Goals/methods

The aim of this research was to determine whether LLMs (i.e., ChatGPT) can bring novel insights into personal reflexive narratives. As autoethnographic reflexivity is an integral part of qualitative research as well as a vital competence of future professionals, this research involved two diverse groups of students (business undergraduates and doctorate researchers). First, the business undergraduates wrote reflection narratives about their cultural identities, whereas the doctorate researchers wrote reflections about their researcher journey in building knowledge of research methodologies in collaboration with GenAI. Second, both groups were asked to carry out thematic analyses of their reflections using the LLMs.

Results

The results indicated that using GenAI support for writing reflections has not effectively benefited knowledge construction. Gen AI provided structured summaries and transformed what was narrated as personal opinions and emotions into more factual accounts. The authenticity and authorship of the narratives were reduced. The answers generated by GenAI were accepted uncritically and reproduced without triangulation, or reflective conceptual judgement.

Conclusions

Although GenAI can provide technical and cognitive support , its use in reflective writing requires pedagogical guidance and critical supervision. Over-reliance on automatically generated texts compromises the quality of reflection and the development of critical thinking, essential in qualitative research and professions. Thus, we argue for the need to integrate GenAI into training practices that emphasize authorship, theoretical triangulation, and ethical discernment.